English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

We aim to tackle a novel vision task called Weakly Supervised Visual RelationDetection (WSVRD) to detect "subject-predicate-object" relations in an imagewith object relation groundtruths available only at the image level. This ismotivated by the fact that it is extremely expensive to label the combinatorialrelations between objects at the instance level. Compared to the extensivelystudied problem, Weakly Supervised Object Detection (WSOD), WSVRD is morechallenging as it needs to examine a large set of regions pairs, which iscomputationally prohibitive and more likely stuck in a local optimal solutionsuch as those involving wrong spatial context. To this end, we present aParallel, Pairwise Region-based, Fully Convolutional Network (PPR-FCN) forWSVRD. It uses a parallel FCN architecture that simultaneously performs pairselection and classification of single regions and region pairs for object andrelation detection, while sharing almost all computation shared over the entireimage. In particular, we propose a novel position-role-sensitive score map withpairwise RoI pooling to efficiently capture the crucial context associated witha pair of objects. We demonstrate the superiority of PPR-FCN over all baselinesin solving the WSVRD challenge by using results of extensive experiments overtwo visual relation benchmarks.

PPR-FCN: Weakly Supervised Visual Relation Detection via Parallel Pairwise R-FCN